HIV is the most common risk factor for the emergence of active tuberculosis (TB) after previously controlled infection (latent infection). Our data suggest that tuberculous granulomas in a single host are independent and dynamic such that clinical outcome is the sum result of these interactions. Our overarching hypothesis is that each lesion bears its own risk of reactivation that is dependent on immunologic, mycobacterial and viral (HIV) interactions. We will focus on two types of asymptomatic, paucibacillary infection commonly disrupted by HIV infection, classic latent infection that occurs after recent infection (i.e., HIV induced reactivaton) and controlled infection after short course treatment for active TB (i.e., HIV driven relapse). These studies will involve the non- human primate model in which M. tuberculosis and SIV (a surrogate to HIV) infection is similar to humans. In aim 1, we will characterize the process of reactivation among CD4 depleted and SIV infected animals with latent infection. We will use bar-coded M. tuberculosis strains, in vivo imaging (PET CT) and molecular methods to detect granuloma-specific growth and killing to examine the lesional conditions that lead to reactivation compared to latent controls in the non-human primate model. We will characterize lesion specific viral load, bacterial burden and immune responses to determine what factors lead to lesional reactivation. In aim 2, we will develop a model for relapse in which asymptomatic, paucibacillary disease occurs after short course treatment for active TB. We will identify bacterial reservoirs of relapse, track bacterial relapse during SIV-Mtb co-infection and determine the lesional risk factors of relapse. In both aims 1 and 2, we will generate a road map of reactivation/relapse, identify the lesions at greatest risk and dissect the immunologic, mycobacterial and viral interactions that contribute to their risk of reactivation/relapse on a lesional level. Many of these findings, especially PET CT characterizations, are likely to be directly translatable to humans. Results of these studies will ultimately lead to better drugs strategies (targeting sites of residual bacteria at high risk of reactivation or relapse) and immun modulating modalities (either as vaccine or supplement to drug treatment).